Update README.md

README.md

@@ -18,6 +18,12 @@ makes a difference between english and English.
 Disclaimer: The team releasing Data2Vec-Text did not write a model card for this model so this model card has been written by
 the Hugging Face team.
 
+## Pre-Training method
+
+![model image](https://raw.githubusercontent.com/patrickvonplaten/scientific_images/master/data2vec.png)
+
+For more information, please take a look at the [official paper](https://arxiv.org/abs/2202.03555).
+
 ## Abstract
 
 *While the general idea of self-supervised learning is identical across modalities, the actual algorithms and objectives differ widely because
@@ -58,9 +64,6 @@ The RoBERTa model was pretrained on the reunion of five datasets:
 
 Together theses datasets weight 160GB of text.
 
-![model image](https://raw.githubusercontent.com/patrickvonplaten/scientific_images/master/data2vec.png)
-
-
 ### BibTeX entry and citation info
 
 ```bibtex